Caution on soundfonts? (Jon Wild question)

> Message: 4
> Date: Wed, 7 Dec 2005 14:26:20 -0500 (EST)
> From: Jon Wild <wild@...>
> Subject: Re: Score and audio files for fifthtone piece (c. 1600?)
>
> On Tue, 6 Dec 2005, Margo Schulter wrote:
>
> > An intriguing keyboard piece possibly from around 1600 appears in a
> > Portuguese manuscript (Coimbra Ms. Mus. 48) with symbols which Hoyle
> > Carpenter has persuasive interpreted as marks raising a note by a meantone
> > diesis or fifthtone of the kind described and embraced by Nicola Vicentino
> > (1555, 1561) and Fabio Colonna (1618). Here are some audio and score
> > files, the later in PostScript, PDF, and abc notation formats:
>
> Hi Margo and thanks, this is very interesting. Three questions for you:
> where is the Hoyle Carpenter publication about the fifthtones? Where does
> Morley talk about the diesis notation (surely not in A Plaine and Easie
> Introduction...)? And, about the tuning: why are some of the G-D fifths
> noticeably more impure than other fifths?
>

Dear Jon,

Please let me explain that I saw your message only after returning from
holidays in Los Angeles, but will attempt a brief reply, since the
questions are most interesting.

In one tuning bibliography, the Carpenter article is cited only as a
manuscript. Johnny Reinhard may have the publication information; I seem
to recall a publication called something like _Acta Musicologica_, and
have tried to locate the article in one university library collection of
that periodical, but so far without success (possibly the issue isn't in
their collection).

The explanation by Morley on the diesis and the enharmonic genus indeed
appears in _A Plain and Easy Introduction to Practical Music_, in the
Annotations to the First Part (the section on the basics of solmisation
and rhythm, etc., which precedes the treatment of counterpoint or discant
and composition), p. 102 in the Alec Harman edition in paperback published
by Norton in 1973. That's ISBN 0-393-00682-4.

As for the beating of some G-D fifths, possible hypotheses would include
a coding error by me when I wrote the ASCII score file for processing by
Aaron Johnson's et_compose. The notation I used is 1200000-et (thousands
of cents as units), so 696578 would be G in a given octave, and 193157
would be D. Here's a link to this source code:

<http://www.bestII.com/~mschulter/Coimbra48.etc>

Other hypotheses might focus on software (including soundfont) issues in
translating this source file into a MIDI file, and from there through
Timidity into an mp3 or ogg file. My guess would be that the MIDI file
produced by Aaron's software is accurate, and that soundfont factors could
produce the beat rates you query. Again, a first guess -- but one
suggested to me by some other technical discussions.

Peace and love,

Margo